We thought it might be helpful to have a way to improve and practice presenting seminars or other public speaking tasks besides chanting in front of a bathroom mirror. We wanted to make an application that would provide a more curated form of feedback to the presenter which helps them improve more quickly.
What it does
It records the visual and audio of an individual speaking and tracks their body movements and facial features to evaluate how engaged they are as speakers and if they are fidgeting too much. It also transcribes their speech to further generate feedback on their pace of speaking for instance and how many filler words they use.
How we built it
We used the Cognitive api from Microsoft's Project Oxford to recognize facial expressions of the speaker and to track the face as well. We used Express for the routing and used d3 to generate some visualizations for the
Accomplishments that we're proud of
Building something that we would use and something that might be helpful to others as well.